As scholars of humble origin, we cannot rise to prominence on our own. We need to have like-minded benefactors in high places and be recommended [into office] through that tie. Only then can we straighten that which is bent and put into motion that which is dormant.1

Yi Kok 李穀 (1298-1351)

*This paper was originally presented at a conference on periodizing Korean history held at the University of Hong Kong. The author acknowledges the generous funding provided by the Academy of Korean Studies Conference and Research Support scheme (AKS-2015-C15) and a matching grant from the School of Languages and Cultures at the University of Hong Kong. Special thanks to Remco Breuker, Allison Van Deventer, Sem Vermeersch, and two anonymous reviewers.

1. 寒門窮巷之士，固不能自達，必有青雲知己爲之援引，乃可以申其屈動其蟄. Yi Kok, “Yŏ tongnyŏn Cho Chungsŏ, Ch’oe Hŏnnap sŏ,” Kajŏng chip (Han’guk munjip ch’onggan ed.) 3.148a.

Javier Cha (javiercha@snu.ac.kr) is an Assistant Professor of East Asian Studies at the College of Liberal Studies, Seoul National University.

Seoul Journal of Korean Studies 32, no. 1 (June 2019): 35–80.© 2019 Kyujanggak Institute for Korean Studies

To Build a Centralizing Regime:Yangban Aristocracy and Medieval Patrimonialism*

Javier Cha

This article examines the ambiguous roles of birth and merit in Korea from around 900 to 1500 CE. Medieval Korea’s yangban aristocracy valued pedigree but did not guarantee heredity. Many illustrious houses failed to retain their status, and the void created opportunities for provincials to try establishing themselves as new and lasting capital-based yangban courtiers. The mesh of patron/client ties and conjugal bonds generated in this social process explain the centralization of authority and the consolidation of the peninsular society after the collapse of first-generation states in the Korea-Manchuria region.

Keywords: Koryŏ, medieval Korea, centralization, sociopolitical history, patrimonialism, patron/client ties, marriage alliance

36 Javier Cha

Introduction

First-generation Korean-Manchurian state-societies were largely structured on the notion of divine rulership. One such example was the bone-rank aristocracy (kolp’um che 骨品制) of Silla 新羅 (–935 CE). The bone-rank system was an unusually rigid and exclusive de jure aristocracy in which a large portion of one’s life was determined by legally defined and legally protected heredity. A detailed set of regulations prescribed the highest office one could attain, the kind and size of house one could build, the type of clothing one could wear, and the districts in which one could reside. The prohibition of marriage across status boundaries reinforced the system’s divisions. Status was a birthright. Limited attempts to introduce meritocratic features into government were met with resistance.

The bone-rank aristocracy was made possible by two far-reaching legacies of the so-called unification wars among early Korea’s Three Kingdoms. First, Silla/Tang allied forces’ victory over Koguryŏ 高句麗 (–668 CE) and Paekche 百濟 (–660 CE) brought Silla under the hegemony of Tang 唐 (618–907). Silla’s sheltered environment acted as a disincentive to institutional innovation or reform, especially efforts that would have sought a wider distribution of authority. Silla’s involvement in coastal trade and cultural exchanges with neighboring communities was for the most part unofficial and associated with powerbrokers and lesser nobles who were disenchanted with the political establishment. Second, Silla and Tang uprooted a large number of conquered Koguryŏ and Paekche nobles and resettled them in ethnic enclaves within the Tang empire.2 The only documented exception was Ansŭng 安勝 (fl. 668–683), who was given true-bone status, and the multiethnic community of Koguryŏ and Malgal 靺鞨 stock attached to him. With almost no major conquered noble houses left in the former territory of the Three Kingdoms, Silla managed to create a rigid aristocracy in the primary and secondary capitals that did not share its authority, status, or privilege with provincials.3 The bone-rank establishment did not even assign surnames, let alone feudal estates or noble titles, to the common folk, regardless of their influence within the local community. The absence of a feudal structure was instrumental for the bone-

2. Chi Paesŏn, Koguryŏ yumin ŭi nara Che wa Tang kŭrigo Silla Parhae Ilbon kyoryusa (Seoul: Hyean, 2012); Chi Paesŏn, Koguryŏ Paekche yumin iyagi (Seoul: Hyean, 2006).

3. Ha Ilsik, Silla chipkwŏn kwallyoje yŏn’gu (Seoul: Hyean, 2006); Chŏn Tŏkchae, Han’guk kodae sahoe ŭi wanggyŏngin kwa chibangmin (Seoul: T’aehaksa, 2002); Chŏn Tŏkchae, Silla wanggyŏng ûi yŏksa (Seoul: Saemunsa, 2009).

To Build a Centralizing Regime 37

rank’s extraordinary exclusivism.Medieval Koreans of Koryŏ 高麗 (918–1392) and the early part of Chosŏn

朝鮮 (1392–1910), for their part, started to build a new sociopolitical order that—peculiarly—both valued birth and demanded that status be earned. Membership in the aristocracy had to be gained through formal appointment in yangban 兩班, or “two branches:” the central government’s civil and military bureaucracy. As a social system, yangban was a de facto aristocracy in which hereditary norms were enforced by social practice. No legislation guaranteed the protection of status or wealth. Status had to be reinforced through civil or military service in the central government, and the gateway to building such a career was the demonstration of one’s talent and capability in the state examinations. Over time, yangban aristocrats added more sources of qualification to the mix, such as membership in prestigious organizations, association with respected intellectual figures, and the memory of an ancestor who had served in civil or military office. A yangban household that failed to fulfill such expectations over several generations effectively lost its right to be treated as yangban.

Why was this the case? Koryŏ was founded by strongmen who toppled the bone-rank order of Silla and participated in another round of unification wars involving the so-called Later Three Kingdoms. By definition, those outsiders were unable to trace their ancestral pedigree to pre-Koryŏ times. With the exception of very few former true-bone nobles, most yangban claimed to have been actively involved in the unification campaigns of the Later Three Kingdoms and to have received their surname from Koryŏ’s founder in return for their merit. The house of Kyŏngwŏn Yi 慶源李氏 claimed to have received the Yi surname when a distant ancestor, an emissary of the Silla court, allegedly accompanied Emperor Xuanzong of Tang 唐玄宗 (r. 712–756) to Sichuan 四川 soon after the outbreak of the An Lushan Rebellion (755–763).4 This fabrication was probably accepted at face value because of the Kyŏngwŏn Yi’s exceptional wealth and power during much of the eleventh century. In most cases, yangban houses were unable to produce verifiable stories of ancestry dated to the times before the tenth-century unification campaigns.

Medieval Korea’s style of government falls neatly into Max Weber’s definition of patrimonial domination in a traditional society. To borrow Julia Adams’ succinct reworking of Max Weber, patrimonialism is characterized by “a style of administration structured by the ruler’s ad hoc, unstable, and always revocable

4. Koryŏsa 95.9a (yŏlchŏn 8, “Yi Chayŏn”) mentions this story as a widely accepted belief.

38 Javier Cha

delegation of powers.”5 In Koryŏ, the central state exercised the authority to retract symbolic and economic rewards at will. The court ennobled hundreds of meritorious princes, courtiers, and office-holding provincials without offering legislated protection of status and wealth. Local strongmen were granted estates and sinecure posts, but at the discretion of personnel representing the will of the center. In principle, titles and estates were not inheritable without the court’s sanction. The medieval aristocracy in Korea was held together strictly by hierarchical relations of personal loyalty. Personalities and houses failed to stay in power for an extended period. Monarchs, princes, noble houses, and illustrious courtiers formed fiercely competitive cliques, and numerous provincials entered the capital in search of opportunities. At the helm, the monarch or the regent and his henchmen wielded tremendous influence and fame; upon falling out of power, entire segments of an aristocratic family risked permanently losing everything their leaders and predecessors had built. The twin solutions of patronage and marriage forged relations of mutual protection with the collective aim of rising to the top and securing advantageous positions by social practice.

Patronage and Marriage in Medieval Korea

Medieval patrimonialism was a mélange of ad hoc creations and institutional experiments. Koryŏ’s search for a working replacement for Silla’s bone-rank aristocracy resulted in a type of government that lacked a clear logic, or, in Weberian terms, a legal-rational bureaucracy with impersonal structures of domination. Despite the wish to create an aristocratic society akin to Silla and to enforce status-based prescriptions in the manner of bone ranks, Koryŏ did not (and could not) guarantee birth rights to monarchs, regents, and courtiers who lacked distinguished ancestors. The degree of power differentiation was low, and the bureaucratic apparatus provided only limited checks and balances.

Yi Sugŏn and John Duncan, two leading medieval historians, have collected the most comprehensive data germane to tracking the changing fortunes of official-producing medieval Korean houses (see Tables 1A: Yi Sugŏn’s List of Major Medieval Houses, 1B: Yi Sugŏn’s List of Medieval Houses, and 1C: John Duncan’s List of Medieval Houses). In collating these data, these historians originally intended to answer the question of whether the history of Korea from the early Koryŏ to the early Chosŏn period should be interpreted as a history of

5. Julia Adams, The Familial State: Ruling Families and Merchant Capitalism in Early Modern Europe (Ithaca, NY: Cornell University Press, 2005), 17.

To Build a Centralizing Regime 39

continuity or change.6 That is, they wanted to know whether the same ruling families that had dominated the high-ranking bureaucratic positions in the tenth and eleventh centuries continued to do so in the fifteenth. Indeed, the medieval houses in these data sets may be arranged and divided into the following types: (1) three to four persistent performers (Type A); (2) twenty to twenty-one prominent early medieval houses that disappear from the political scene by the fourteenth century (Type B); (3) twenty-seven to thirty newcomers (Type C) who enter the medieval court during the eras of military regency (1170–1270) or Mongol rule (1270-1351); (4) and those that seem to follow irregular patterns (Type D). These data sets should not be analyzed as representative samples of medieval Korean society. While it is striking that only three to four known medieval houses consistently produced officials from the beginning to the end of Koryŏ, the question about continuity or change is not a productive way of gaining insight into the workings of medieval patrimonialism. These lists of medieval houses are valuable for what the individual houses in the aforementioned four divisions qualitatively represent, such as perished houses.

The twenty or twenty-one perished houses identified through the rearrangement of the data collected by Yi Sugŏn and John Duncan include Ch’ungju Yu 忠州劉氏, Yŏngch’ŏn Hwangbo 永川皇甫氏, Kwangyang Kim 光陽金氏, and Kyŏngwŏn Yi (cf. Table 2B: The Background of Early and Middle Koryŏ Queens). In the tenth and eleventh centuries, these noble houses were the byword for power and wealth, and they monopolized the supply of queens and regents to the throne. Other houses joined the court later, during the activist reforms in the late eleventh and early twelfth centuries: e.g., the Haeju Ch’oe 海 州崔氏, Namp’yŏng Mun 南平文氏, and Chŏngan Im 定安任氏, which produced the likes of Ch’oe Sach’u 崔思諏 (1034–1115), Mun Kongmi 文公美 (d. 1137), and Im Wŏnhu 任元厚 (1089–1156). A grandson of Ch’oe Ch’ung 崔冲 (984–1068), Ch’oe Sach’u was deeply involved in the activist programs of Kings Sukchong 肅宗 (1095–1105) and Yejong 睿宗 (1105–1122); Mun Kongmi and his brothers were a family from a destitute background that rose to prominence as part of the Han Anin faction in the 1110s and 1120s; Im Wŏnhu, the offspring of a prominent institutional reformer, survived the political infighting of the early 1120s and enjoyed a spectacular career along with his two brothers. In the post–Han Anin era, Im Wŏnhu betrothed his daughter to King Injong 仁宗 (r. 1122–1146) to become one of the most powerful noblemen in Koryŏ. All these

6. John Duncan, “The Social Background to the Founding of the Chosŏn Dynasty: Change or Continuity?” Journal of Korean Studies 6 (1988–9): 39–79; John Duncan, The Origins of the Chosŏn Dynasty (Seattle: University of Washington Press, 2000); Yi Sugŏn, Han’guk chungse sahoesa yŏn’gu (Seoul: Ilchogak, 1984).

40 Javier Cha

houses shared two features: they enjoyed unrestrained power during times of political success but struggled to pass down their status, privilege, and authority to posterity. In fact, most of their descendants completely disappeared from Koryŏ’s political scene after three to four generations. What happened?

Medieval Koreans regarded lifelong nobles and the royal kin as courtiers of effectively equal footing. Irregular succession, regicide, and usurpation took place time and again, and there was no guarantee of a smooth transition of authority to the reigning monarch’s posterities (see Table 2A: Royal Succession in Early and Middle Koryŏ). As a result, the king, paralleling the behavior of his courtiers, prudently built networks of marriage and patronage to strengthen his authority and safeguard the passage of the throne to his designated heir. He found it difficult, however, to bypass the looming presence of powerful nobles, who were at once his most dependable associates and immediate threats.

Some traditional historians recognized this paradox and expressed their estimations at length, though modern historians tend to dismiss such writings as mere projections of “Confucian” or moralistic historiography. Yi Chehyŏn 李齊 賢 (1287–1367), a fourteenth-century Confucian scholar-official, left behind a thought-provoking and succinct appraisal of King Munjong 文宗 (r. 1046–1083) that enumerated the points of success and failure during the reign of the monarch responsible for launching activist reforms. In the paragraphs that follow, I will delineate the workings of the Koryŏ court as a commentary on Yi Chehyŏn’s opinion piece.

Kings Hyŏnjong 顯宗 [r. 1009–1031], Tŏkchong 德宗 [r. 1031–1034], Chŏngjong 靖宗 [r. 1034–1046], and Munjong passed the throne from father to son or between brothers. Those eighty years may be called an age of splendor.7

Yi Chehyŏn defined the eighty years leading up to the late eleventh-century activist reforms as “an age of splendor” solely on the basis of regular succession. As a hereditary monarchy, the Koryŏ Dynasty determined the order of succession on the basis of a twofold formula. The preferred arrangement was patrilineal succession by primogeniture, but at times the throne was passed to a brother if the heir was deemed too young to oversee the realm effectively. The ambiguity of this formula divided the court into interest groups advocating for a contender to succeed to the throne. The patriarch of an influential house established conjugal bonds with the claimant and upon successful enthronement served as the regent. Beginning in the early years of the dynasty, the fight for the throne

7. 顯德靖文父作子述兄終弟及首尾，幾八十年可謂盛矣. Koryŏsa chŏryo 5.43a (King Munjong 37, 1083).

To Build a Centralizing Regime 41

surfaced as a source of political fragmentation and intrigue. King T’aejo 太祖 (r. 918–943), the dynastic founder, upon establishing marriage relations with more than twenty local strongmen to pacify the provinces, designated his elder son as the official heir to the throne. However, “fearing that [the heir] would not be able to accede the throne due to the humble background of Consort O, his mother,”8 King T’aejo placed the crown prince under the protection of Pak Surhŭi 朴述熙 (d. 945), a strongman from Hyesŏng County 槥城郡 (present-day Tangjin 唐津) who is described in his official biography as a man “of lionhearted character who enjoyed devouring meats of all sorts including toads, frogs, crickets, and ants.”9 The plan succeeded to a degree, and King Hyejong 惠宗 (r. 943–945), the designated heir, ascended to the throne. However, his reign was curtailed with the purge of Pak Surhŭi, his protector, and his own death under suspicious circumstances after less than two years of rule. Thereafter, the throne changed hands between his half-brothers, and for nearly thirty years Koryŏ experienced frequent conflicts among supporter groups struggling to crown their royal candidates.

This game of thrones prompted the monarch and his fraternal contenders to forge conjugal bonds with one another and with influential houses. Over generations, Koryŏ’s princes took intermarriage with close kin and cognatic recognition of relations and status as the norm, which bewildered the later medieval Confucian historians. One anonymous official historian claimed: “King T’aejo was determined to transform our customs after the models of antiquity. However, people customarily followed the native practice of betrothing sons to their own daughters, who deceptively would assume their maternal surname as their own. Their offspring regarded this as the household norm without questioning its aberrance. Alas!”10 Whether marriage between close kin and the cognatic assignment of surname were “native practices” (t’osŭp 土習) is questionable.11 Intraclan marriage probably took place for pragmatic reasons. King Hyejong, for example, “consigned his elder princess to [his brother] So to strengthen his authority. The princess assumed the surname Hwangbo of her mother’s clan. Since then, everyone who practiced intraclan marriage has

8. 以其母吳氏側微恐不得立. Koryŏsa 92.12a (yŏlchŏn 5, “Pak Surhŭi”).9. 述熙性勇，敢嗜啗肉，雖蟾蜍螻蟻皆食. Koryŏsa 92.12a (yŏlchŏn 5, “Pak Surhŭi”).10. 太祖法古有志化俗，然狃於土習，以子聘女諱稱外姓，其子孫視爲家法，而不之怪，惜哉！ Koryŏsa 88.1a-b.

11. Martina Deuchler also questions whether marriage customs in Koryŏ adhered to any particular set of prescriptive regulations. Martina Deuchler, The Confucian Transformation of Korea: A Study of Society and Ideology (Cambridge, MA: Harvard University Asia Center, 1991), 71–73.

42 Javier Cha

adopted their maternal surname.”12 This marriage between an uncle and his niece was the fruit of the reigning monarch’s effort to placate his fraternal rival. Upon marriage, the princess adopted her maternal surname, though in fact it was not Hwangbo, but Im, the surname of her maternal grandfather Im Hŭi 林曦 (tenth century).13 The adoption of affinal surnames, usually the mother’s surname or the father’s maternal grandmother’s surname, among married daughters of royal kin may have represented the princess’s or queen’s affiliation with a particular faction in the court society’s hostile milieu.

That is not to suggest that exogamy was the convention in Koryŏ society. Broadly speaking, “Marriages with close cognates were customary,”14 as Martina Deuchler, No Myŏngho, and Yi Sugŏn have demonstrated in detail, but the collapse of the bone-rank aristocracy lifted the legal restrictions and social taboos against marriage across status boundaries.15 In Koryŏ, marriage and patronage took place on the basis of pragmatic considerations. Numerous studies that aimed to discover patterns in Koryŏ society’s marriage norms proved futile, probably because the medieval Korean elite observed few, if any, prescriptive regulations in partner selection. On several occasions in the medieval era, reform-minded monarchs called for the prohibition of marriage between relatives belonging to the mourning circle. Most such efforts failed for nearly five centuries, and they continued to emerge as a subject of court discussion as late as the fifteenth century.16 Nonetheless, Koryŏ’s aristocrats could not afford to limit their affinal connections to clan members and other nobles of comparable standing. Their desire to compete in the political arena and bequeath their privileged position to posterity forced them to seek out promising outsiders.

The period between 1009 and 1083, shared by the reigns of Kings Hyŏnjong, Tŏkchong, Chŏngjong, and Munjong, marked a transition away from the tenth century’s muddled politics. The succession and marriage patterns of kings

12. 王...以長公主妻昭用强其勢，公主從母姓稱皇甫氏，後凡取同姓者皆諱稱外家之姓. Koryŏsa chŏryo 2.2a (King Hyejong 2, 945).

13. The anonymous historian in Koryŏsa chŏryo wrote “Hwangbo” in error. See Koryŏsa 88.9b (hubi 1, Kyŏnghwagung Puin Im ssi). Her mother was Queen Ŭihwa, née Im of Chinju.

14. Deuchler, The Confucian Transformation of Korea, 65.

15. In Silla, the bone-rank institution strictly prohibited marriage across established status boundaries. See Peter Lee et al., Sources of Korean Tradition: From Early Times through the Sixteenth Century (New York: Columbia University Press, 2000), 26–27; Deuchler, The Confucian Transformation of Korea; No Myŏngho, “Koryŏ sahoe ŭi yangch’ŭk chŏk ch’insok chojik yŏn’gu” (Ph.D. dissertation, Seoul National University, 1988); and Yi Sugŏn, Han’guk ŭi sŏngssi wa chokpo (Seoul: Sŏul taehakkyo ch’ulpansa, 2003).

16. See, for example, Koryŏsa 84.39a-b (chi 38, hyŏngpŏp 1, kongsik, Kanbi (in 1308 and 1367)) and Sŏngjong sillok 9 [1478/07/19].

To Build a Centralizing Regime 43

between 936 and 1170 reveal an intriguing shift (see Tables 2A and 2B). After King Hyejong’s suspicious death in 945, the throne passed to his half-brother with the backing of a military faction. His brother’s reign lasted a mere four years. King Kwangjong 光宗 (r. 949–975), his fraternal rival, succeeded King Chŏngjong 定宗 (r. 945–949), and the court society enjoyed relative stability during the twenty-six years of his rule. For thirty-two years after the end of Kwangjong’s reign in 975, Koryŏ experienced a period of grave instability. The throne changed hands four times, including twice by irregular succession and once by regicide, while the kings defended the realm against a series of devastating Khitan invasions. King Hyŏnjong’s accession in 1009 marked a new beginning. Eleventh-century monarchs turned to court bureaucrats for support while aiming to rid themselves of dependence on tenth-century high nobles, including the Yŏngch’ŏn Hwangbo, Chŏngju Yu 貞州柳氏, and Ch’ungju Yu. The early Koryŏ kings eventually succeeded (see Table 2B: The Background of Early and Middle Koryŏ Queens).17 They devised a hybrid strategy that combined—with no apparently consistent patterns—marriage with agnatic relatives and non-agnatic courtiers, including both affinal relatives and those with no blood relations. The eleventh-century marriage network among the Kyŏngju Kim 慶州金氏, Ich’ŏn Sŏ 利川徐氏, Tanju Han 湍州韓氏, and Kyŏngwŏn Yi allowed peaceful transfers of power for reasons implied in Yi Chehyŏn’s second point.

King Munjong was diligent and frugal. He employed men of talent and wisdom and lightened penal sentences out of affection for his people. He venerated learning and respected the elders. Noble titles and attire were not conferred on scoundrels; power and authority did not transfer to their kin. Even among his close kin, he did not reward those without merit; even among his most favored men, he punished those with faults.18

Kings Munjong, Sukchong, and Yejong aimed to build an activist monarchy that operated through institutions, free of interventions from royal kin and power-wielding courtiers. The new regime downplayed personal ties of reciprocity and promoted loyalty to the dynastic state as an adumbration of Samhan 三韓 civilization. Ch’oe Sajŏn 崔思全 (1067–1139) admonished King Injong: “The Samhan is the Samhan of the Samhan and does not stop at being Your Majesty’s Samhan. King T’aejo, our former lord, worked laboriously to

17. The data collected by John Duncan and Yi Sugŏn strongly suggest their disappearance from the court bureaucracy. See Tables 1B and 1C.

18. 而文宗躬勤節儉，進用賢才，愛民恤刑，崇學敬老，名器不假於匪人，威權不移於近．雖戚里之親，無功不賞，左右之愛，有罪必罰. Koryŏsa chŏryo 5.43a (King Munjong 37, 1083).

44 Javier Cha

arrive at where we are. I beg Your Majesty not to be negligent of this task.”19 In this discourse of abstract bureaucratic order, the head of state assumed the responsibility of operating his court on meritocratic principles, valuing experience, talent, and acts of distinction. Those procedures were expected to be handled with minimal attempts by court personalities to sway political decisions in their favor, with the understanding that impartial proceedings would earn the compliance and trust of the people without the need to resort to harsh penal laws.

The state examination institution was the emblem of this impartial order. This elaborate procedure of staff recruitment demanded that the offspring of courtiers prove their worth and provided a crucial gateway of opportunity for qualified provincials to seek bureaucratic advancement at the center. By the reign of King Munjong, one century had passed since the institution’s intro- duction in 958, and Koryŏ’s courtiers had by then accepted it as the most respected means of entry into office. In 986, the Koryŏ court had awarded merely fifty degrees after thirteen examination sessions and had faced difficulty in encouraging provincial candidates to relocate to the capital for scholarly training. King Sŏngjong 成宗 (r. 981–997) opted to provide a long-term incentive to foster the growth of education:

Despite my shameful lack of virtue, I sincerely venerate scholars and intend to restore the mores of the Duke of Zhou and Confucius. I hope to effect the model of Yao and Shun, nurturing scholars in schools and recruiting them via examin- ations. Now, some of you, students who entered the capital from the provinces, yearn for your people at home. I hereby permit you to decide whether you will stay or leave at will. All of you mark my words—you are not to abandon your study. I award 1400 bolts of silk to 207 of you who elected to return home, and the fifty-three of you remaining in the capital will be granted 106 hats and 265 sŏk of rice.20

The situation changed considerably in the following century, as demonstrated by the aggregation of Koryŏ state examination data in twenty-year intervals (see Chart 1).21 After experiencing a steady decline until 1040, the examination

19. 三韓者，三韓之三韓也，非止陛下之三韓也．先君太祖，勤勞以致，請勿忽之. Koryŏ myojimyŏng chipsŏng, “Ch’oe Sajŏn myojimyŏng,” no. 33. The translation follows Remco Breuker with minor changes. See Remco Breuker, Establishing a Pluralist Society in Medieval Korea, 918–1170 (Leiden: Brill, 2010), 34.

20. 朕素慙薄德，尙切崇儒，欲興周孔之風，冀致唐虞之理，庠序以養之，科目以取之，今諸州所上學生，慮有思鄕之人，皆令從便去留，汝等祗稟予言，勿墜其業，其歸寧學生二百七人，可賜布一千四百匹，留京學生五十三人，亦賜幞頭一百六枚，米二百六十五碩. Koryŏsa 28.25a (chi 28, “Kukhak”).21. Unlike Chosŏn, Koryŏ examinations were not held on a regular triennial schedule, but

To Build a Centralizing Regime 45

system steadily increased the number of degrees granted, and the state operated twelve prestigious provincial academies to attract candidates.22 The Koryŏ state passed about two hundred candidates every twenty years during institutional reforms between 1040 and 1100. From 1100 to the demise of the Ch’oe regency in 1258, that figure soared to an average of four hundred students, as did the number of students enrolled in the National Academy (Kukchagam 國子 監 or Kukhak 國學). In 1130, the court, facing a situation that would have been absurd to King Sŏngjong in 986, cut expenditures in public education upon identifying the National Academy as a major revenue drain.23 The students

irregularly upon demand. Therefore, these data do not need to be aggregated at an interval that is a multiple of three years.

22. Pak Ch’ansu, Koryŏ sidae kyoyuk chedosa yŏn’gu (Seoul: Kyŏngin munhwasa, 2001).

23. Koryŏsa chŏryo 9.49b-50a. (King Injong 8, 1130).

Chart 1. Civil Examination Degrees Conferred in Koryŏ(at 20-year intervals)

46 Javier Cha

launched a collective protest and successfully reversed the decision.By the eleventh century, thousands of provincials were drawn to the allure of

the center. The activist reforms in the late eleventh century inspired these provincials to compete for a bureaucratic career, although the availability of such positions was severely limited. At the same time, activist monarchs improved the size and quality of the standing army, enlarged the tax base, and streamlined the bureaucratic procedures. The total number of staff was kept low to reduce expenditures, which in turn further improved the state’s fiscal health and reduced the collective influence of the nobility. Hundreds of students flocked into the examination halls, and thousands of men of valor joined the military. In 1107, more than six hundred candidates took the civil examinations, and only twenty-seven received degrees.24 The Koryŏ court subsequently restricted participation to three times in a lifetime to cope with this intractable demand.25 Likewise, military institutions experienced unprecedented growth. According to Xu Jing’s observation, as many as thirty thousand soldiers were stationed on active duty in Kaesŏng 開城 in 1123.26

Through this combination of examinations and patronage, the Koryŏ court attracted a large number of provincials to the center. Many personalities with biographies listed in The Dynastic History of Koryŏ were reportedly men of “obscure” or “humble” background, without explicit mention of their legally ascribed status. For example, Mun Kongmi of Namp’yŏng came from “a humble household,”27 Kwŏn Sup’yŏng 權守平 (d. 1250) was described as “a man from Andong, of obscure past and unknown house pedigree,”28 and Kim Chŏngsun 金正純 (1086–1145) was described simply as someone who “arose from destitute conditions.”29 All of those men ascended the social ladder in the mid-Koryŏ period. Mun Kongmi had no notable officeholding ancestors before his father, Mun Ik 文翼 (fl. 1098–1102). Kwŏn Sup’yŏng was probably a local functionary who sought gradual promotion into the court bureaucracy by means of a sinecure post. Kim Chŏngsun, whose ancestral seat was not

24. “Kŭpche pangbang kyosŏ,” Tongmunsŏn 23.

25. The entry for the eighth reign year of King Ŭijong (1154) in Koryŏsa 73.10a (chi 27, “kwamok”) states: “Our country’s regulation stipulates that the participation in the examinations be limited to no more than three times.” 國制…就試者例不過三.26. Xu Jing, “Longhu Zhongmeng jun,” Gaoli tujing, juan 11.

27. 家世單寒. Koryŏsa 125.1b (yŏlchŏn 38, Mun Kong’in).28. 安東人, 跡微不知其族譜. Koryŏsa 102.18b (yŏlchŏn 15, Kwŏn Sup’yŏng).29. Koryŏsa 98.24b (yŏlchŏn 11, Kim Chŏngsun): “Arising from destitution, Kim Chŏngsun was a chivalrous man who valued resoluteness and excelled in archery and horseback riding.” 正純起自寒素，尙氣任俠，善射御．

To Build a Centralizing Regime 47

mentioned in the primary sources, may have been a commoner and was apparently “a chivalrous man who valued resoluteness and excelled in archery and horseback riding.”30 His gifts in the military arts helped him earn merit in the northern pacification campaigns, thereby securing his status as a bona fide court bureaucrat.

Activist monarchs benefited immensely from the inundation of provincials who needed royal benefaction to boost their career prospects. Meritocracy in Koryŏ ultimately served the interests of the king or regent, and the various avenues of entry into officialdom were neither a product of goodwill nor a result of impartial institutions administered with legal rationality. The exami- nation system operated on a set of regulations that designated the career path of candidates based on their subject and excluded certain undesired segments of the population from official candidacy. In practice, the king or regent frequently made exceptions. Unusually, the aforementioned Ch’oe Sajŏn advanced to civil officialdom with a medical examination degree. He relocated to Kaesŏng from Cheju Island as a palace doctor, maneuvered shrewdly in court politics, and rose to a second-rank chancellery post.31 During an examination procedure, the court appointed a Chief Examiner (chigonggŏ 知貢擧) to oversee the selection of worthy candidates, but upon the conclusion of each session, the king personally assumed the responsibility of final evaluation.32 The monarch’s participation in the proceedings, which occurred regularly until 1110, was more than a symbolic gesture; it communicated his intention to provide personal benefaction to the candidates from that point forward.33

Numerous examination candidates who by legislation should have been disqualified advanced to the court bureaucracy by royal prerogative. In 1043, the court prohibited unfilial sons, the lowborn, and their descendants from partaking in the examinations34 and, under institutional reforms, expanded the scope of the prohibition to everyone who lacked official record of his ancestry in household registration documents and to the offspring of incestuous unions.35 No Chun 魯準 (fl. eleventh century) was born to parents belonging to intersecting mourning circles. He attained the high-level chinsa 進士 degree nonetheless, but the court objected to his birth from an endogamous union when he sought

30. Ibid.

31. Koryŏ myojimyŏng chipsŏng, “Ch’oe Sajŏn myojimyŏng,” no. 33.

32. Koryŏsa 73.2a-47a (chi 27, “kwamok”).

33. Ibid.

34. Koryŏsa 73.2a-47a (chi 27, “kwamok”).

35. This also appears in Koryŏsa 73.2a-47a (chi 27, “kwamok”).

48 Javier Cha

promotion. Ch’oe Sŏk 崔奭 (fl. 1051–1088), the Minister of Personnel, enacted a move to terminate his bureaucratic career, against which King Munjong opined: “In recruiting talent, we need not strictly adhere to ordinary regulations. Perhaps he could be awarded a bureaucratic post to open a path into court membership.”36 King Munjong and Ch’oe Sŏk disagreed, but eventually reached a compromise, following the latter’s recommendation and rationale: “Ordered household leads to proper governance of the state. [No] Chun’s father had an indecent marriage that disturbs the cardinal human relationships. However, our ongoing veneration of Confucian methods established the recruitment of scholars as the priority. It may be fitting to assign him a lower-rank duty.”37 In another case, Chŏng Mun 鄭文 (d. 1106), whose maternal grandfather was from a lowborn district, attained a civil examination degree to enter officialdom. His lowborn link surfaced as a point of debate only when he was recommended for a censorial position. King Sukchong agreed with the arguments launched against appointing him as a censor, but promoted him to other upper-tier civil bureaucratic posts. Chŏng Mun served as an important contributor to activist reforms. In 1105, he served as the crown prince’s tutor and Minister of Rites, and subsequently completed a successful diplomatic mission to the Northern Song 北宋 (960–1127), where he acquired a large number of books.

Besides the king, Koryŏ’s courtiers also engaged in the practice of patronage. In the early eleventh century, Yi Chujwa 李周佐 (d. 1040), described in the official biography as “a man of Kyŏngju, who came from a humble household but was a bright child,” was spotted by Yi Sŏnggong 李成功 (fl. ca. 997–1039), then the District Regent of the Eastern Capital (tonggyŏng yusu 東京留守), who brought Yi Chujwa to Kaesŏng.38 At the capital, Yi Chujwa enrolled in the National Academy at the recommendation of his patron and passed the civil examinations sometime between 1000 and 1009, leading to a bureaucratic career that reached the senior third rank. Later in the same century, Kwak Sang 郭尚 (1034–1106) experienced rapid promotion in social status from a functionary to a courtier: “Kwak Sang, styled Wŏllo, was a man of Ch’ŏngju. Originally a minor functionary, he climbed up the social ladder by currying

36. 選用人才，當不拘常局，可與諸進士，並授官秩，以通朝籍. Koryŏsa 95.28a-b (yŏlchŏn 8, Mun Chŏng).

37. 家齊然後國治. 隼父不正婚禮，瀆亂人倫，然方今崇尙儒術，用士是急，宜降授階職. Koryŏ sa 95.28b (yŏlchŏn 8, Mun Chŏng).

38. Koryŏsa 94.39b (yŏlchŏn 7, Yi Chujwa): “Yi Chujwa, a man of Kyŏngju, came from a humble household but was a bright child. While stationed as District Regent of the Eastern Capital, Yi Sŏnggong, Vice Director of the Royal Secretariat, recognized his talent upon first sight. He brought [Yi Chujwa] along with him to the capital when he was recalled from his post.” 李周佐慶州人，家世單微，幼聰悟，左僕射李成功留守東京，一見器之，及還携至京，使隸國學．

To Build a Centralizing Regime 49

favor and served King Sŏnjong at National Residence.”39 Of course, “currying favor” is simply a way of describing in a negative light Kwak’s ability to seek and establish connections with the right people in positions of power.

Whether within civil or military officialdom, a successful bureaucratic career in the medieval court society required the establishment of strong bonds with influential patrons, preferably an aristocratic courtier or a top-tier military official. Permanent resettlement in the dynastic center, in contrast, necessitated marriage alliances with courtiers of equal or higher status. The gradual enforce- ment and adoption of consanguinity laws in later medieval times curbed the widespread practice of kinship and regional endogamy, and two centuries of hegemonic rule under the Ch’oe dictatorship and Mongol domination consid- erably disturbed the foothold of the early medieval aristocracy in the Koryŏ court. Between circa 1170 and 1350, many local functionaries and men of obscure origin progressively filled the bureaucracy alongside the surviving members of the old aristocracy. Those who hedged their bets prudently in this political game of patronage and marriage became the interchangeably civil and military courtiers known as the yangban, who preserved their privileged status through inclusion in a cohesive web of relinked affines. The Andong Kwŏn 安東 權氏 coalition was one among many such intellectual and marriage networks of courtiers.

Marriage and Patronage under Mongol Rule

The Mongol rule of Korea from 1258/1270 to 1351 was a turning point in the making of the yangban aristocracy. Going back to the list of medieval houses collated by Yi Sugŏn and John Duncan, this discussion concerns the twenty-seven to thirty newcomers grouped under Type C (see Table 1B and Table 1C). Like their early and middle Koryŏ counterparts, the majority of members of late Koryŏ houses gained employment in the central government’s civil or military service—that is, yangban—via patronage and sought to secure their relocation to the capital through marriage with other yangban. The dynamic changed in three ways. First, the Mongol leaders in Daidu 大都 (Khanbaliq 汗八里; present-day Beijing), the capital of Yuan 元 (1271–1368), replaced the civil and military regents as the head patrons of yangban courtiers. Second, examination degrees declined in value and popularity for about sixty years after the Mongol conquest

39. Koryŏsa 97.8a (yŏlchŏn 10, Kwak Sang). 郭尚字元老，淸州人．起小吏夤緣攀附，事宣宗于國原邸．

50 Javier Cha

of Koryŏ; however, this trend was reversed soon thereafter (see Chart 1). Examples are abundant. Consider the Haengju Ki 幸州奇氏 and P’yŏngyang

Cho 平壤趙氏, who were notable beneficiaries of Mongol patronage. The Ki originally worked as Ch’oe House retainers and gained political power when one of the daughters was married to Emperor Shundi of Yuan 元順帝 (r. 1333–1368; Toghon Temür 妥懽帖睦爾). Cho In’gyu 趙仁規 (1237–1308), the progenitor of the P’yŏngyang Cho, was a low-ranking functionary or sinecure official in a small district near P’yŏngyang. His political success came from his proficiency in Mongolian, and after working as the Koryŏ king’s translator, he rapidly moved up to become an ennobled courtier with patronage and marriage ties in Daidu. Others, such as the Ch’angnyŏng Sŏng 昌寧成氏 and Kwangju Yi 廣州李氏, were local strongmen who entered the capital by collective examination performance. Sŏng Yŏwan 成汝完 (1309–1397) entered office in 1336 upon attaining a civil examination degree and worked his way up in the court bureaucracy as an active participant in King Kongmin’s 恭愍王 (r. 1351–1374) reforms. When Yi Sŏnggye 李成桂 (1335–1408) launched the dynastic coup, he opposed the move and retreated to the countryside in protest. However, Sŏng Sŏngnin 成石璘 (1338–1423), his son, joined the northeasterners’ cause and was given a merit subject privilege. In the case of the Kwangju Yi, the first try at the court bureaucracy happened with Yi Chip 李集 (1314–1387), who passed the civil examinations in 1368 but held an official post for no more than three years. Their conversion into a bona fide yangban house had to wait until Yi Inson 李仁孫 (1395–1463) supported the Usurpation of 1454; his clansmen attained examination degrees in large numbers over a short period soon thereafter.

Many of these newcomers originated in the south. As the data collected by Yi Sugŏn and Kang Ŭn’gyŏng suggest, Mongol rule devastated the power base of courtiers and functionaries with their ancestral seat in the northwest (see Table 4B: Local Strongmen After Mongol Rule).40 This “devastation” involved

40. Yi Sugŏn, Han’guk ui sŏngssi wa chokpo (Seoul: Sŏul taehakkyo ch’ulpansa, 2003), 179–186; Yi Sugŏn, “Koryŏ-Chosŏn sidae chibae seryŏk pyŏnch’ŏn ŭi che sigi,” in Han’guksa sidae kubullon, edited by Ch’a Hasun et al. (Seoul: Sohwa, 1995), 241–247; Kang Ŭn’gyŏng, Koryŏ sidae hojang ch’ŭng yŏn’gu (Seoul: Hyean, 2002), 254. Both Yi Sugŏn and Kang Ŭn’gyŏng identify the unusual-ly high proportion of “perished surnames” (mangsŏng 亡姓) in P’yŏngan Province as an anomaly but do not advance a plausible explanation for this observation. Interestingly, Yi Sugŏn presents two records from The Annals of the Chosŏn Court, dated to 1413 and 1466, that, in hindsight, attribute the unusually low number of “established surnames” (t’osŏng 土姓) in P’yŏngan region to warfare and defection to the Ming. These sources do not draw a direct link to the Yuan, however. This aspect will need further research. 惟東西二界，曩在前朝，累經兵亂，州郡騷然，田疇荒穢，移民入鎭，以供守禦. T’aejong sillok 26 [1413/11/5]; 平安道西隣大明，而人皆入鎭，本無土姓．且近日勒令墾田，人多憚行. Sejo sillok 40 [1466/11/2].

To Build a Centralizing Regime 51

the near-total elimination of those established surnames. Table 4B replicates Kang Ŭn’gyŏng’s province-level aggregation of “established” and “perished” surnames based on her interpretation of the various categories that appear in the gazetteer. The subsequent three tables are my breakdown of three of those provinces. This simple exercise of aggregation and ratio calculation reveals an astonishing pattern. The southernmost provinces of Chŏlla and Kyŏngsang remained unscathed, while the central provinces of Hwanghae, Kyŏnggi, and Kangwŏn saw forty to forty-five percent of established functionary surnames disappear. Established surnames in Kyŏngsang Province collectively comprised nearly eighty-three percent of all of the surnames represented, whereas in Kyŏnggi the equivalent figure was fifty-eight percent. The figure in Kyŏnggi paled in comparison with areas further north. In P’yŏngan Province, the gazetteer identifies a total of 452 surnames. Of those, only twenty-seven, or six percent, are listed as established surnames. A staggering ninety-four percent, or 425, were outsiders, and 404 out of those 425 had moved in, probably as a result of the post-Kongmin Koryŏ and early Chosŏn states’ resettlement policy.

One of the most successful southern clans that laid the foundation for their lasting yangban status was the Andong Kwŏn. The Kwŏns claimed descent from the progenitor Kwŏn Haeng 權幸/行 (fl. early tenth century), a strongman from the southeastern county of Koch’ang 古昌郡. During the civil war in the early tenth century, he initially opposed King T’aejo 太祖, but he later reversed his position.41 For his cooperation, Kwŏn Haeng received investiture as a “merit subject” and saw his home county promoted to Andong prefecture 安東府. In 930, the inchoate Koryŏ state sent him back to his ancestral seat and appointed him Taesang 大相, a local functionary position in the fourth rank.42 For ten generations, the Kwŏns resided in Andong as an “established surname” (t’osŏng 土姓) and preserved their status as local notables through a balance of kinship endogamy and marriage with other local clans of similar pedigree.43 Starting from the tenth generation, two segments achieved courtier status and forged a cohesive network of marriage alliances consisting exclusively of courtier, royal, and imperial lineages. Patronage and marriage played decisive roles in those two segments’ status promotion and long-term preservation of their elevated social position.44

41. No Myŏngho, Koryŏ kukka wa chiptan ŭisik: chawi kongdongch’e, Samguk yumin, Samhan ilt’ong, Haedong ch’ŏnja ŭi ch’ŏnha (Seoul: Sŏul taehakkyo ch’ulpansa, 2009), 15; Tongguk yŏji sungnam, Andong, inmul.

42. Koryŏsa 57.27b (chi 11, chiriji 2, Andong-bu).

43. Yi Sugŏn, Han’guk ŭi sŏngssi wa chokpo, 234.

44. In this regard, my interpretation differs from those of Pak Yongun and Yi Chŏngho. Pak

52 Javier Cha

In early and middle Koryŏ, the Kwŏn generally remained content with their position as local functionaries and did not seek bureaucratic careers in Kaesŏng. That is, with the exception of one documented personality. Kwŏn Chŏk 權適 (1094–1147; not shown in the network graph) passed the literary examinations in 1112 at the age of nineteen. However, the sudden death of his father before the final stage of the examination obstructed his entry into the court bureaucracy. Upon completing the funerary rite, he caught the attention of Yi Chahyŏn 李資玄 (1061–1125) at Mañjuśrı̄ Monastery (Munsusa 文殊寺) near present-day Ch’unch’ŏn 春川. This encounter proved a turning point for Kwŏn’s career, as Yi Chahyŏn was of the illustrious Kyŏngwŏn Yi house and the younger brother of the regent Yi Chagyŏm 李資謙 (d. 1126). In 1115, when the Northern Song court invited Koryŏ’s bureaucrats to receive training in the Imperial Academy (Guozijian 國子監), Kwŏn Chŏk was granted the exceptional privilege of being sent to Kaifeng 開封 along with four other promising students, a privilege he likely owed to Yi Chahyŏn’s considerable influence on the recommendation procedure. As contrived as this practice of patronage may appear, Yi Chahyŏn precociously evaluated the ability and talent of this youthful academician. His circle of immediate agnatic and affinal network, after all, produced the most respected men of letters in Kaesŏng at the time. In China, Kwŏn Chŏk impressed Emperor Huizong 徽宗 (r. 1100–1126), attained the prestigious jinshi degree 進士, received an honorary bureaucratic title, and returned to Koryŏ in 1117 along with a team of emissaries led by Yi Chahyŏn’s brother Yi Charyang 李資諒 (d. 1123).45 Kwŏn Chŏk subsequently enjoyed three decades of a successful bureaucratic career in the middle and upper-middle ranks until his death in 1147 while serving as Assistant Master of the National Academy (kukchagam cheju 國子監祭酒; junior third rank).46

Kwŏn Chŏk experienced a meteoric rise in status: starting as a local notable,

Yongun accepts at face value the Andong Kwŏn’s claim, which is likely a fifteenth-century fabrication, that their distant ancestor was a Kim-surnamed aristocrat in Silla. To support his view that Koryŏ was an aristocratic society (kwijok sahoe 貴族社會), he downplays the importance of the Kwŏns’ relocation from Andong to the capital during the era of military regency. See Pak Yongun, “Andong Kwŏn ssi ŭi sarye rŭl t’onghae pon Koryŏ sahoe ŭi iltanmyŏn: Sŏnghwabo rŭl ch’amgo ro hayŏ,” Yŏksa kyoyuk 94 (2006): 35–86. Yi Chŏngho examines the rise of the Kwŏn Chungsi/Kwŏn Sup’yŏng segment to prominence during the Mongol domination period. He credits this rise to a new economic foundation that the Andong Kwŏn built using the official stipend and downplays the importance of extortion and Yuan patronage. Yi Chŏngho, “Koryŏ hugi Andong Kwŏn ssi kamun ŭi kyŏngje chŏk kiban: Kwŏn Chungsi/Kwŏn Sup’yŏng kyeyŏl ŭl chungsim ŭro,” Han’guk sahakpo 21 (2005): 333–366.

45. Koryŏsa 14.22a (sega 14, King Yejong 12, 1117).

46. Koryŏsa 17.5b-6a (sega 17, King Injong 19, 1141); Koryŏ myojimyŏng chipsŏng, “Kwŏn Chŏk myojimyŏng,” no. 46.

To Build a Centralizing Regime 53

he became a middle-level civil official who visited China. However, his family lasted only a single generation in the capital, as did the family of Kwŏn Chŏk’s great-grandfather. According to Kwŏn Chŏk’s epitaph, his great-grandfather Kwŏn Kyunhan 權均漢 (dates unknown) served as a low military official in the junior ninth rank, while his grandfather and father held honorary titles attributed posthumously. After a brief stint in the capital, they probably returned to their ancestral role as hereditary functionaries in Andong. Kwŏn Chŏk’s literary talent and participation in the civil examinations opened the door for this provincial man to enter a bureaucratic career in Kaesŏng, but his success thereafter depended heavily on the benefaction of the influential courtier Yi Chahyŏn. The Andong Kwŏn’s second documented attempt at courtier status turned out to be ephemeral. Following the purge of the Kyŏngwŏn Yi in 1126, Kwŏn Chŏk failed to secure ties with a new influential patron and competed with other established courtiers. His five sons failed to live up to their father’s achievement as an eminent academician; at the time of Kwŏn Chŏk’s death, Kwŏn Taebang 權大方 (fl. 1148) was a student at the National Academy and along with Kwŏn Tollye 權敦禮 (fl. 1148) held a sinecure position; Kwŏn Palchin 權發真 (fl. 1148) and Kwŏn Tonsin 權敦信 (fl. 1148) became Buddhist monks; and Kwŏn Chŏk’s youngest son was still a toddler. None of his descendants engaged in activities worthy of documentation, and the editors of the 1476 genealogy did not register Kwŏn Chŏk and his children as members of the Andong Kwŏn descent group.47

Descendants of the ninth-generation progenitor Kwŏn Iyŏ 權利輿 (dates unknown) drafted the 1476 edition of the Andong Kwŏn genealogy, which for the initial eight generations recorded only the direct line of ancestors leading to Kwŏn Iyŏ and suppressed the entry of the extended branches, citing the paucity of sources. Kwŏn Chŏk’s father Kwŏn Tŏgyŏ 權德輿 (fl. circa 1094) may have belonged to the same generation as Kwŏn Iyŏ, but this connection cannot be established empirically.

Regardless of the precise line of descent connecting Kwŏn Chŏk and his possible uncle Kwŏn Iyŏ, more Kwŏns from Andong pressed to become a courtier household in Kaesŏng. Three segments of Kwŏn Iyŏ’s line attained bureaucratic positions and enjoyed varying degrees of success at relocating to the capital and solidifying their courtier status. Kwŏn Iyŏ and his sons remained in Andong as functionaries; status differentiation began to occur in the

47. The 1476 edition of the Andong Kwŏn genealogy records the initial eight generations between the progenitor Kwŏn Haeng and Kwŏn Iyŏ by their names only and with the extended branches suppressed. Whether this decision was made for the sake of simplification or due to a dearth of empirical evidence is unclear.

54 Javier Cha

subsequent generation. The line of Kwŏn T’ong 權通 (dates unknown) remained as hereditary functionaries in Andong for an additional four documented generations. Kwŏn Ch’ehwa 權棣和 (dates unknown), of another segment, succeeded in receiving a sinecure military post in the capital, but his patrilineal line ended prematurely with his two sons. One son became a Buddhist monk, and the other son, with unknown social standing, produced three daughters. Two of his daughters married the local functionaries Kwŏn Yangjun 權良俊 (dates unknown) and Kim Igyŏn 金利堅 (dates unknown), both probably of Andong, in what may be evidence of kinship and regional endogamy. The above examples, from Kwŏn Chŏk to Kwŏn Iyŏ, demonstrate the Andong Kwŏns’ eagerness to relocate to Kaesŏng as well as the difficulties they encountered in this process.

Kwŏn Sup’yŏng succeeded in permanently relocating his family to the capital, though he did not live to see his offspring grow into one of the most illustrious descent groups in early modern Korea. The official biographical sketch in The Dynastic History of Koryŏ paints Kwŏn Sup’yŏng as a frugal man of obscure ancestry who lived in destitute conditions.48 At the height of the Ch’oe regency, he entered as a military official of the junior ninth rank and patiently worked his way up in the military branch of the bureaucracy. By the time he left office, he had served as deputy head at the Bureau of Military Affairs (ch’umirwŏn pusa 樞密院副使), a prestigious position in the senior third rank. No details survive regarding the life of his son, Kwŏn Wi 權違 (dates unknown), except that he worked as a Hallim Academician (Hallim haksa 翰林學士). This generation marked the Andong Kwŏn’s permanent change from a military to a civil officeholding yangban house.

Subsequently, the unambitious Kwŏn Tan 權㫜 (1228–1311) entered the picture. As a youth, Kwŏn Tan wished to lead a life in seclusion but decided against it at the urging of his father, who apparently exhausted the family’s finite resources to provide for Tan’s education. Tan had a knack for literary composition and passed the civil examinations upon the recommendation of the Chancellor Yu Kyŏng 劉璥 (1211–1289), who told him: “You have training in the literary arts; you ought not be doing administrative tasks.”49 His successful career as a local magistrate, civil bureaucrat, and state examiner ended with voluntary early retirement. A pious follower of Buddhism who had followed a vegetarian diet for forty years, Kwŏn Tan finally, in his late years, fulfilled his wish to become a monk despite the repeated protests of Kwŏn Pu 權溥 (1262–

48. Koryŏsa 102.18b-19b (yŏlchŏn 15, Kwŏn Sup’yŏng).

49. 子有文學不宜爲吏. Koryŏsa 107.12b (sega 20, Kwŏn Tan).

To Build a Centralizing Regime 55

1346), his ambitious heir.Kwŏn Pu played a pivotal role in establishing the Andong Kwŏn as the most

illustrious and recognizable descent group in medieval and early modern Korean history. In the Chosŏn Dynasty, his offspring, along with members of a secondary branch, would produce more than three hundred civil examination degree holders, second in rank only to the royal family.50 According to a demographic survey from 2015, Andong Kwŏn is the eleventh largest choronym in South Korea, with a total population of nearly 700,000.51 The origins of this spectacular rise in number, wealth, and prestige can be traced back to 1279, when Kwŏn Pu passed the civil examinations at the tender age of seventeen and started his bureaucratic career. He stayed in office for a staggering sixty-seven years, forty years of which were in chancellery positions, until his death in 1346 at the age of eighty-four.52

Under Kwŏn Pu’s patriarchy, the Andong Kwŏn witnessed the ennoblement of five sons and four sons-in-law, maximizing the advantageous aspects of Mongol rule. Wang Hu 王煦 (1296–1349), born Kwŏn Chae 權載, entered office at the recommendation of his brother Chun 權準 (1280–1352). Both King Ch’ungsŏn 忠宣王 (r. 1298–1313) and King Ch’ungsuk 忠肅王 (first reign 1313–1330; second reign 1332–1339) took a liking to Kwŏn Chae. The former king conferred on him the royal surname, adopting Kwŏn Chae as his godson Wang Hu, and the latter king ennobled him as the Lord of Kyerim (Kyerim Puwŏn’gun 雞林府院君). In 1314, the Yuan court deposed King Ch’ungsŏn and recalled him to Daidu. Wang Hu followed his king to the Yuan court, during which time he built a solid network of patrons and received the Mongol name Tuohuan 脫歡. After returning to Kaesŏng in 1325, he quickly rose to the rank of State Councilor and endeavored to tackle the widespread problems of extortion and land grabbing. Ironically, many of the culprits were his own family members and their political associates.

During the era of Mongol rule, families backed by powerful patrons used their clout to accumulate wealth by illicit means. In the final evaluation of Kwŏn Pu, The Dynastic History of Koryŏ states: “As a sociable character, he held the position in charge of appointing personnel for a long time, building his wealth by selling offices. The people at the time thought of him as a far cry from his father’s frugal life.”53 Other families took a more aggressive stance, the most

50. Aggregated from the Munkwa pangmok 文科榜目 data.51. Data aggregated from the Korean Statistical Information Service: http://kosis.kr/.

52. Koryŏsa 107.12b-27b (yŏlchŏn 20, Kwŏn Tan).

53. 爲人無圭角，久典銓衡，鬻爵營産，時人以爲：『視其父㫜之淸懸，遠也.』 Koryŏsa 107.13a (yŏlchŏn

56 Javier Cha

common act being the seizure of estates and slaves. No Ch’aek 盧頙 (d. 1356) was said to have “a rapacious disposition that enjoyed seizing others’ slaves.”54 Ki Samman 奇三萬 (d. 1347), a relative of Ki Ch’ŏl 奇轍 (d. 1356), “relied on his political influence to unlawfully take lands belonging to others.”55 He died in the midst of an interrogation headed by Wang Hu. A brother of Wang Hu, Kwŏn Kyŏm 權謙 (d. 1356), eagerly sought Mongol patronage early in his career, for which he became a myriarch (wanhu 萬戶) and was able to marry his daughter to the Yuan crown prince.56 In 1356, Kwŏn Kyŏm colluded with Ki Ch’ŏl and No Ch’aek to depose King Kongmin because of conflicting views regarding the diminishing Mongol hegemony. The plot failed, resulting in the execution of Kwŏn Kyŏm, Ki Ch’ŏl, and No Ch’aek along with their henchmen and the court’s seizure of their property. King Kongmin cited the illicit seizure of land and property as one of the reasons for their purge. One edict read: “Ki Ch’ŏl and Kwŏn Kyŏm have marriage ties to the imperial house. Drawing on the power of this influence, they defied the rule of law and seized the people’s land and possessions and engaged in wrongdoing at will.”57

King Kongmin’s edict in no way overstated the scale of their illicit activities. Ki Wŏn 奇轅, a brother of Ki Ch’ŏl, led an extravagant lifestyle that was described as follows: “Once, Ki Wŏn gathered his clansmen to host a banquet for his mother using extremely luxurious vessels, utensils, and delicacies. Observers thought of this as a rare sight in Korea.”58 However, Ki Wŏn’s lifestyle was not as extraordinary as this account implies. Kwŏn Chun 權準 (1281-1352), a loyal supporter of King Kongmin who died before his brother committed treason in 1356, owned a luxurious compound. King Ch’ungsuk was said to have toured his residence in admiration at one point and exclaimed: “This is well beyond my own ability.”59 As for how Kwŏn Chun managed to build this luxurious compound that astonished even his king, the official biographical sketch in The Dynastic History of Koryŏ gives an explanation that by now must sound like a

20, Kwŏn Tan).

54. 頙性貪，好奪人臧獲. Koryŏsa 131.22a (yŏlchŏn 44, No Ch’aek).55. 轍族弟三萬亦倚勢恣行不法，奪人土田. Koryŏsa 131.13b (yŏlchŏn 44, Ki Ch’ŏl).56. 福安府院君權謙，如元納女于皇太子，元拜大府監大監. Koryŏsa 38.12a (sega 38, King Kongmin 11, 1352).

57. 奇轍、權謙，連姻帝室，依勢作威，不畏紀綱，奪占田民，恣行非義. Koryŏsa 39.26b-27a (sega 39, King Kongmin 8, 1359).

58. 轅嘗會宗族，宴其母，器皿珍羞，窮極侈麗，見者以爲：『東韓以來罕有也.』 Koryŏsa 131.15a (yŏlchŏn 44, Ki Ch’ŏl).

59. 忠肅嘗移御準第，周觀屋宇之美，歎曰：『非寡躬所敢當也.』 Koryŏsa 107.16a (yŏlchŏn 20, Kwŏn Tan).

To Build a Centralizing Regime 57

common trope: “He accumulated his massive wealth by relying on his political influence to seize lands and take bribes.”60 Kwŏn Ko 權皐 (1294–1379), the final Andong Kwŏn of this generation to be discussed, provides a slightly different story. He had a property dispute with his son, Kwŏn Kan 權侃 (fl. 14th century), that led to a litigation battle. When his son did not show up for the hearing, Kwŏn Ko “kicked Kan’s wife out of anger, causing a miscarriage as well as her death.”61 During the official investigation of this incident, an unnamed figure apparently testified: “Kwŏn Ko is a rapacious and violent character. He who kicked his son’s wife to death is no father. Kan, who disobeyed his father’s will, is no son either.”62 The early Chosŏn editors of Koryŏ records did not deliberately insert these episodes to vilify the degenerate conditions of the preceding dynasty. Indeed, these editors included the offspring and intellectual heirs of the Andong Kwŏn. While the death of a daughter-in-law was clearly regarded as a tragedy, and some details might have been adjusted or amplified, violent clashes occurred rather frequently in Koryŏ under Mongol rule. In this fiercely competitive environment, the new breed of yangban courtiers ruthlessly built power and wealth, a process that involved land grabbing, extortion, and sometimes kidnapping, rape, and murder. The Andong Kwŏn survived this turbulent late Koryŏ era and reaped the benefits of this gain for centuries afterward.

Conclusion

The foregoing discussion is a new way of thinking about the ambiguous roles of birth and merit in medieval Korean society, which have been treated as a paradox in the established historiography and periodization schemas. Under the influence of modernization theory, historians of Korea have approached the medieval era using the dichotomy of feudal aristocracy and bureaucratic centralization. This reworking of Max Weber implicitly defines the latter as a more evolved “modern” form of government that operates on the meritocratic qualities of impersonal institutions, which in preindustrial East Asian societies were purportedly provided by the Confucian tradition. Thus, a heated debate in modern historiography has been about the point at which Korean society transitioned from patrimonial domination to legal-rational forms of governance.

60. 倚勢奪土田納賄賂以致鉅富. Koryŏsa 107.16b (yŏlchŏn 20, Kwŏn Tan).61. 怒蹴侃妻墮胎死. Koryŏsa 107.18a (yŏlchŏn 20, Kwŏn Tan).62. 皐本貪殘人也，蹴殺子婦，非父也，侃忤父意非子也. Koryŏsa 107.16a (yŏlchŏn 20, Kwŏn Tan).

58 Javier Cha

There is a remarkable lack of consensus on whether this Weberian social change took place in Koryŏ, in Chosŏn, or at all in preindustrial Korea. Some historians opine that Buddhism was the state religion of Koryŏ, and that Koryŏ marked a gradual shift from the “native” customs and practices of the Three Kingdoms era to the Neo-Confucian world of Chosŏn. However, Remco Breuker’s revision paints Koryŏ as a world in which “native” and “Buddhist” ideas and practices coexisted alongside Confucianism and the pluralistic notion of Samhan unifi- cation.63 The extent of the bureaucratic centralization in Koryŏ has remained a point of dispute as well. The political system of Koryŏ is described as a combi- nation of “aristocratic” and “bureaucratic” elements because in staff recruitment, the Koryŏ state adopted the civil service examinations in a manner that valued candidates’ pedigree. With respect to official appointment, the Koryŏ state did appoint magistrates, but only to about thirty percent of local districts.

In the 1970s, Pak Ch’anghŭi and Kim Ŭigyu proposed that the implementation of the state examination institution distinguished Koryŏ from Silla’s putative autocratic despotism.64 However, Yi Kibaek and Pak Yongun questioned the extent to which the Koryŏ state’s bureaucratic apparatus penetrated into the lower levels of administration.65 Their empirical research laid bare the medieval Korean state’s remarkably limited capacity to dispatch magistrates to the localities—the abovementioned thirty-percent figure—and the endurance of “feudal” characteristics in local administration, the economy, and land tenure well into the later stages of the dynasty. Yi and Pak held the opinion that the “bureaucratic” stage of Korea’s historical development had to await the coming of Neo-Confucian rule in Chosŏn.

In North America, James Palais and John Duncan laid out another perspective in which the focus was the persistence of “aristocratic” elements throughout Silla, Koryŏ, and Chosŏn, despite the process of Confucian bureaucratization. This thesis of “aristocratic/bureaucratic balance” challenged the claims of “progress” and “modern” elements made by some South Korean historians. The making of this framework began in Palais’ 1975 study, Politics and Policy in

63. Breuker, Establishing a Pluralist Society.

64. Kim Ŭigyu, “Koryŏ kwanin sahoe ŭi sŏngkyŏk e taehan sigo,” Yŏksa hakpo 58 (1973): 61–76; Kim Ŭigyu, Koryŏ sahoe ŭi kwijokche sŏl kwa kwallyoje non (Seoul: Chisik sanŏpsa, 1985); Pak Ch’anghŭi, “Koryŏ sidae kwallyoje e taehan koch’al,” Yŏksa hakpo 59 (1973): 35–59.

65. Yi Kibaek, Koryŏ kwijok sahoe ŭi hyŏngsŏng (Seoul: Ilchogak, 1990); Pak Yong’un, Koryŏ sahoe wa munbŏl kwijok kamun (Seoul: Kyŏngin munhwasa, 2003); James Palais, “Land Tenure in Korea: Tenth to Twelfth Centuries,” Journal of Korean Studies 4 (1982–1983): 73–205; James Palais, “Confucianism and the Aristocratic/Bureaucratic Balance in Korea,” Harvard Journal of Asiatic Studies 44, no. 2 (1984): 427–468.

To Build a Centralizing Regime 59

Traditional Korea, in which he meticulously examined the recalcitrance of landed aristocratic elements in implementing state-directed modernization efforts in the late nineteenth century. In the 1980s and 1990s, Palais extended the application of his thesis to the medieval and early modern eras.66 This effort began with a series of critical evaluations of canonical studies in private land ownership, social mobility, and Confucian intellectuals and institutions. John Duncan, who aimed to reaffirm Palais’s thesis of aristocratic/bureaucratic balance, gathered a wealth of empirical data to argue for the enduring dominance of early Koryŏ “aristocratic” households into the subsequent Chosŏn Dynasty. In Duncan’s version of aristocratic/bureaucratic balance, Koryŏ’s elite consisted of “a core group of descent groups [that] persisted from the beginning to end”67 but also experienced some “minor accretions and deletions…over the centuries.”68 Duncan attributed this social change to the civil service examinations and the beginnings of bureaucratic governance in Korea.

This article has proposed a third way of understanding the social dynamics of medieval Korea: as a push/pull relationship between the center and the provinces. The implementation of “bureaucratic” elements was a way to become an “aristocratic” society in the absence of legislated means of status protection. After the collapse of bone-rank order, Koryŏ assigned surnames to local strong- men who had collaborated with the regime during the unification campaign. A few among them wielded tremendous power and prestige as ennobled houses in the capital, but the influence of these illustrious houses rarely lasted more than four generations. Monarchs and regents to the throne who led the patrimonial aristocracy exercised nearly unrestricted authority, but only in select parts of the realm and for short periods. The central state was connected to the provinces via strong and vertical relationships of coercion and loyalty, but with limited institutional reach into local communities. Over time, however, the yangban sociopolitical order managed to establish long webs of interdependence between the capital and the provinces and increased state centralization to levels previously unknown in the Korea-Manchuria region.

Koryŏ was a volatile environment. In Kaesŏng, the struggle for power among the monarch and noble houses prompted members of the political establishment to recruit outsiders as a means of enhancing their political leverage, and conjugal bonds were accepted as the most trusted practice of solidifying such

66. James B. Palais, Politics and Policy in Traditional Korea (Cambridge, MA: Harvard University Press, 1975).

67. Duncan, The Origins of the Chosŏn Dynasty, 151.

68. Duncan, The Origins of the Chosŏn Dynasty, 151.

60 Javier Cha

patron/client ties. On the flip side, instability in the establishment created a wealth of opportunities for outsiders who aspired to higher social positions. Koryŏ’s transformation into an “aristocratic” society happened by social practice, rather than by prescription, to address a challenge that was shared by the dynastic house, noble houses, and provincial newcomers. This practice was closely entwined with the processes associated with “bureaucratic” centralization, such as staff recruitment by civil examinations. The adoption of Sinitic offices and the civil service examinations was a consequence of this fierce competition for power at the center. With the exception of those with Silla’s true-bone or head-rank ancestry, Koryŏ’s social elite consisted of those with no pedigree in pre-Koryŏ times. In this environment, protection appointment of offices and the recruitment of new retainers from a pool of exam degree holders via patronage and marriage provided a means of self-perpetuation.

The formation of the medieval aristocracy consisted of two major phases. Initially, the push/pull relationship between the center and provinces generated strong centripetal forces. This push/pull mechanism had an unstable center that experienced rapid turnover but forced provincials to establish ties because the state exercised the authority to revoke privileges. After the period of Mongol rule, as shown in the case of the Andong Kwŏn and their associates, this arrangement began to stabilize to the point where the life chances of provincials entering the capital substantially diminished. The coalitions of yangban forged robust ties of mutual protection via patronage and marriage, and the success of this practice transformed the center into a stable environment with a pool of capital-based courtier houses that performed exceptionally in the state examin- ations. A related social change during this period was the adoption of Confucian marriage practices, which prohibited marriage with kin within six degrees of consanguinity. As a consequence, prominent courtier houses in the later medieval period did not enjoy such a high concentration of political power as did their early medieval counterparts, such as the Kyŏngwŏn Yi. The marriage restrictions forced the yangban to diversify the pool of their marriage partners, which had the inadvertent effect of promoting the lineage’s long-term survival and contributing to the growing stability of the center.

Over time, the coalescence of the central establishment discouraged provincials from considering the risky endeavor of advancement and relocation to the capital. With the center enjoying a sociopolitical stability that it had not known for centuries previously, the latecomers to the post-Mongol rush to relocate to the capital had to devise creative new ways of marking their distinction at the local level. Neo-Confucian learning was an intrinsic part of this new local culture, and so was the formation of various organizations to

To Build a Centralizing Regime 61

protect local interests. The middle of the sixteenth century saw the beginnings of the early modern localizing order, which from the medieval standpoint was the perfection of an aristocratic order that would have been the envy of early Koryŏ monarchs. From this long-term perspective, the early modern localizing order was an unlegislated status system that was made possible by a stable center and a constellation of provincials who pursued local interests while remaining loyal to the state.

62 Javier Cha

Table 1A. Yi Sugŏn’s List of Major Medieval Houses69

(number of bureaucrats at third rank and above)

Medieval House 本貫姓氏Early Koryŏ(918–1170)

Military Regency

(1170–1270)

Mongol Domination(1270–1351)

Prominence in the 15th century

Kyŏngju Kim 慶州 金 27 8 4 Y

Inju Yi(Kyŏngwŏn Yi)

仁州 李(慶源 李)

21 5 2 Y

Haeju Ch’oe 海州 崔 17 4 2 Y

P’yŏngsan Pak 平山 朴 16 0 0 N

Suwŏn Ch’oe 水原 崔 14 0 0 N

Changdan Han 長湍 韓 12 2 0 N

Chŏngju Yu 貞州 柳 11 0 0 N

Ich’ŏn Sŏ 利川 徐 9 1 0 Y

Kwangyang Kim 光陽 金 9 0 0 N

Namyang Hong 南陽 洪 8 4 25 Y

Kyŏngju Ch’oe 慶州 崔 8 1 1 N

Kangnŭng Kim 江陵 金 8 0 0 N

Chunghwa Kim 中和 金 8 0 0 N

Chŏngan Im 定安 任 7 5 2 N

Ch’ŏngju Yi 清州 李 7 3 3 N

Paekch’ŏn Cho 白川 趙 7 2 1 Y

Ch’ungju Yu 忠州 劉 7 0 0 N

Hwangju Hwangbo

黃州 皇甫 7 0 0 N

Pongsŏng Yŏm (P’aju Yŏm)

峰城 廉(坡州 廉)

6 1 6 N

Wŏnju Wŏn 原州 元 6 0 11 Y

Ch’ŏngju Kim 清州 金 6 0 0 N

P’yŏngsan Yu 平山 瘐 6 0 0 N

Yŏhŭng Min 驪興 閔 5 3 8 Y

Chuksan Pak 竹山 朴 5 3 5 Y

69. Yi Sugŏn, “Koryŏ-Chosŏn sidae chibae seryŏk,” 239.

To Build a Centralizing Regime 63

Medieval House 本貫姓氏Early Koryŏ(918–1170)

Military Regency

(1170–1270)

Mongol Domination(1270–1351)

Prominence in the 15th century

P’ap’yŏng Yun 坡平 尹 5 2 5 Y

Yŏnggwang Kim 靈光 金 5 0 0 N

Namp’yŏng Mun 南平 文 4 3 1 N

Yangch’ŏn Hŏ 陽川 許 4 1 12 Y

Pup’yŏng Yi 富平 李 4 0 0 N

Naju Na 羅州 羅 3 1 5 Y

Ch’ŏrwŏn Ch’oe 鐵原 崔 2 6 4 N

Andong Kwŏn 安東 權 2 3 14 Y

Hapch’ŏn Yi 陜川 李 2 2 4 N

Hoengsŏng Cho 橫城 趙 2 2 2 N

Kyŏngju Yi 慶州 李 2 1 10 Y

Sŏngju Yi 星洲 李 2 1 9 Y

Ch’ungju Chi 忠州 池 2 0 5 N

Andong Kim 安東 金 1 3 13 Y

Chŏnju Ch’oe 全州 崔 1 3 5 N

Chŏnju Yu 全州 柳 1 2 0 N

Sangju Kim 尚州 金 1 0 4 Y

Chinju Kang 晉州 姜 1 0 4 Y

Kangnŭng Ch’oe 江陵 崔 1 1 2 Y

Mokch’ŏn U 木川 于 0 4 0 N

Ch’ŏngju Kyŏng 清州 慶 0 2 3 Y

Ubong Ch’oe 牛峰 崔 0 5 0 N

P’yŏnggang Ch’ae 平康 蔡 0 2 5 N

Munhwa Yu 文化 柳 0 3 4 Y

Chinju Yu 晉州 柳 0 1 4 N

Haengju Ki 幸州 奇 0 6 3 Y

Ŏnyang Kim 彥陽 金 0 3 8 N

Yŏju Yi 驪州 李 0 2 2 N

Table 1A. (Continued)

64 Javier Cha

Medieval House 本貫姓氏Early Koryŏ(918–1170)

Military Regency

(1170–1270)

Mongol Domination(1270–1351)

Prominence in the 15th century

Kwangju Kim(Kwangsan Kim)

光州 金(光山 金)

0 4 19 Y

Hansan Yi 韓山 李 0 2 2 Y

Sunch’ang Sŏl 淳昌 薛 0 1 4 N

Ch’ŏngju Han 清州 韓 0 0 9 Y

Yŏhŭng An 順興 安 0 0 6 Y

Miryang Pak 密陽 朴 0 0 5 Y

Kosŏng Yi 固城 李 0 0 5 N

Ch’ŏngju Chŏng 清州 鄭 0 0 4 N

P’yŏngyang Cho 平壤 趙 0 0 10 Y

Tanyang U 丹陽 禹 0 0 9 Y

Yŏngsan Sin 靈山 辛 0 0 5 Y

Haep’yŏng Yun 海平 尹 0 1 6 N

Table 1A. (Continued)

To Build a Centralizing Regime 65

Table 1B. Yi Sugŏn’s List of Major Medieval Houses70

(houses that produced three or more bureaucrats at third rank and above)

Medieval House 本貫姓氏Early Koryŏ(918–1170)

Military Regency

(1170–1270)

Mongol Domination(1270–1351)

Type

Ch’ŏngju Yi 清州 李 Y Y Y A

Chuksan Pak 竹山 朴 Y Y Y

Kyŏngju Kim 慶州 金 Y Y Y

Namyang Hong 南陽 洪 Y Y Y

Chŏngan Im 定安 任 Y Y N B

Haeju Ch’oe 海州 崔 Y Y N

Kyŏngwŏn Yi 慶源 李 Y Y N

Namp’yŏng Mun 南平 文 Y Y N

Ch’ŏngju Kim 清州 金 Y N N

Ch’ungju Yu 忠州 劉 Y N N

Chandan Han 長湍 韓 Y N N

Chŏngju Yu 貞州 柳 Y N N

Chunghwa Kim 中和 金 Y N N

Hwangju Hwangbo 黃州 皇甫 Y N N

Ich’ŏn Sŏ 利川 徐 Y N N

Kangnŭng Kim 江陵 金 Y N N

Kwangyang Kim 光陽 金 Y N N

Kyŏngju Ch’oe 慶州 崔 Y N N

P’yŏngsan Pak 平山 朴 Y N N

P’yŏngsan Yu 平山 瘐 Y N N

Paekch’ŏn Cho 白川 趙 Y N N

Pup’yŏng Yi 富平 李 Y N N

Suwŏn Ch’oe 水原 崔 Y N N

Yŏnggwang Kim 靈光 金 Y N N

Andong Kim 安東 金 N Y Y C

Andong Kwŏn 安東 權 N Y Y

70. Yi Sugŏn, “Koryŏ-Chosŏn sidae chibae seryŏk,” 239.

66 Javier Cha

Medieval House 本貫姓氏Early Koryŏ(918–1170)

Military Regency

(1170–1270)

Mongol Domination(1270–1351)

Type

Ch’ŏrwŏn Ch’oe 鐵原 崔 N Y Y

Chŏnju Ch’oe 全州 崔 N Y Y

Haengju Ki 幸州 奇 N Y Y

Kwangju Kim(Kwangsan Kim)

光州 金(光山 金)

N Y Y

Munhwa Yu 文化 柳 N Y Y

Ŏnyang Kim 彥陽 金 N Y Y

Ch’ŏngju Chŏng 清州 鄭 N N Y

Ch’ŏngju Han 清州 韓 N N Y

Ch’ŏngju Kyŏng 清州 慶 N N Y

Ch’ungju Chi 忠州 池 N N Y

Chinju Kang 晉州 姜 N N Y

Chinju Yu 晉州 柳 N N Y

Haep’yŏng Yun 海平 尹 N N Y

Hapch’ŏn Yi 陜川 李 N N Y

Kosŏng Yi 固城 李 N N Y

Kyŏngju Yi 慶州 李 N N Y

Miryang Pak 密陽 朴 N N Y

P’yŏnggang Ch’ae 平康 蔡 N N Y

P’yŏngyang Cho 平壤 趙 N N Y

Sangju Kim 尚州 金 N N Y

Sŏngju Yi 星洲 李 N N Y

Sunch’ang Sŏl 淳昌 薛 N N Y

Tanyang U 丹陽 禹 N N Y

Yŏhŭng An 順興 安 N N Y

Yŏngsan Sin 靈山 辛 N N Y

Mokch’ŏn U 木川 于 N Y N D

Ubong Ch’oe 牛峰 崔 N Y N

Table 1B. (Continued)

To Build a Centralizing Regime 67

Medieval House 本貫姓氏Early Koryŏ(918–1170)

Military Regency

(1170–1270)

Mongol Domination(1270–1351)

Type

Naju Na 羅州 羅 Y N Y

P’ap’yŏng Yun 坡平 尹 Y N Y

Pongsŏng Yŏm(P’aju Yŏm)

峰城 廉 (坡州 廉)

Y N Y

Wŏnju Wŏn 原州 元 Y N Y

Yangch’ŏn Hŏ 陽川 許 Y N Y

Chŏnju Yu 全州 柳 N N N

Hansan Yi 韓山 李 N N N

Hoengsŏng Cho 橫城 趙 N N N

Kangnŭng Ch’oe 江陵 崔 N N N

Yŏju Yi 驪州 李 N N N

Table 1B. (Continued)

68 Javier Cha

Table 1C. John Duncan’s List of Medieval Courtiers, circa 981–140571

Medieval House 本貫姓氏Early Koryŏ(981–1146)

Late Koryŏ(unspecified)

Early Chosŏn(1392–1405)

Type

Hwangnyŏ Min 黃驪 閔 Y Y Y A

Kyŏngju Kim 慶州 金 Y Y Y

P’ap’yŏng Yun 坡平 尹 Y Y Y

Kaesŏng Wang 開城 王 Y Y N B

Kongam Hŏ 公巖 許 Y Y N

Kyŏngju Ch’oe 慶州 崔 Y Y N

Ansan Kim 安山 金 Y N N

Ch’ogye Chŏng 草溪 鄭 Y N N

Ch’ungju Yu 忠州 劉 Y N N

Chŏngan Im 定安 任 Y N N

Chŏngju Yu 淸州 劉 Y N N

Ich’ŏn Sŏ 利川 徐 Y N N

Kangnŭng Kim 江陵 金 Y N N

Kwangyang Kim 光陽 金 Y N N

Namp’yŏng Mun 南平 文 Y N N

Okku Im 沃溝 林 Y N N

P’yŏngsan Pak 平山 朴 Y N N

P’yŏngsan Yu 平山 庾 Y N N

Pongju Chi 鳳州 智 Y N N

Suju Ch’oe 水州 崔 Y N N

Suju Yi 樹州 李 Y N N

Tanju Han 湍州 韓 Y N N

Yŏngch’ŏn Hwangbo

永川 皇甫 Y N N

Yŏnggwang Kim 靈光 金 Y N N

Andong Kim 安東 金 N Y Y C

Andong Kwŏn 安東權 N Y Y

71. Based on Duncan, The Origins of the Chosŏn Dynasty, 57.

To Build a Centralizing Regime 69

Medieval House 本貫姓氏Early Koryŏ(981–1146)

Late Koryŏ(unspecified)

Early Chosŏn(1392–1405)

Type

Ch’ŏngju Han 清州 韓 N Y Y

Chŏnju Ch’oe 全州 崔 N Y Y

Chuksan Pak 竹山 朴 N Y Y

Kwangsan Kim 光山 金 N Y Y

Kyŏngju Yi 慶州 李 N Y Y

Munhwa Yu 文化 柳 N Y Y

Namyang Hong 南陽 洪 N Y Y

P’yŏngyang Cho 平壤 趙 N Y Y

Sŏngju Yi 星洲 李 N Y Y

Sunhŭng An 順興 安 N Y Y

Tanyang U 丹陽 禹 N Y Y

Andong Chang 安東 張 N N Y

Ch’angnyŏng Sŏng 昌寧 成 N N Y

Ch’ŏngju Chŏng 清州 鄭 N N Y

Ch’ŏngju Kyŏng 清州 慶 N N Y

Ch’ŏngsŏng Sim 青松 沈 N N Y

Chinju Kang 晉州 姜 N N Y

Chŏnju Yi 全州 李 N N Y

Hanyang Cho 漢陽 趙 N N Y

Kosŏng Yi 固城 李 N N Y

Miryang Pak 密陽 朴 N N Y

Paekch’ŏn Cho 白川 趙 N N Y

Ponghwa Chŏng 奉化 鄭 N N Y

Sŏngju Pae 星州 裵 N N Y

Ŭiryŏng Nam 宜寧 南 N N Y

Yŏnan Kim 延安 金 N N Y

Yŏngch’on Hwangbo

永川 皇甫 N N Y

Yŏngil Chŏng 迎日 鄭 N N Y

Table 1C. (Continued)

70 Javier Cha

Medieval House 本貫姓氏Early Koryŏ(981–1146)

Late Koryŏ(unspecified)

Early Chosŏn(1392–1405)

Type

Ch’ŏngju Kwak 清州 郭 Y N Y D

Ch’ŏngju Yi 淸州 李 Y N Y

Haeju Ch’oe 海州 崔 Y N Y

Kyŏngwŏn Yi 慶源 李 Y N Y

Tongnae Chŏng 東萊 鄭 Y N Y

Onyang Kim 溫陽 金 N Y N

P’yŏngyang Ch’ae 平壤 蔡 N Y N

Wŏnju Wŏn 原州 元 N Y N

Table 1C. (Continued)

To Build a Centralizing Regime 71

Table 2A. Royal Succession in Early and Middle Koryŏ, 918–1270

Predecessor Successor Year Relationship Process Trend

T’aejo Hyejong 943 elder son hereditary (backed by a regent) unstable

Hyejong Chŏngjong 945 half-brother fraternal competition(possible regicide)

Chŏngjong Kwangjong 949 brother fraternal competition

Kwangjong Kyŏngjong 975 elder son hereditary (backed by a regent)

Kyŏngjong Sŏngjong 981 cousin elected caretaker

Sŏngjong Mokchong 997 nephew returned to rightful heir

Mokchong Hyŏnjong 1009 cousin/uncle regicide (backed by a regent)

Hyŏnjong Tŏkchong 1031 elder son hereditary relatively stable

Tŏkchong Chŏngjong 1034 brother fraternal

Chŏngjong Munjong 1046 half-brother fraternal

Munjong Sunjong 1083 son hereditary

Sunjong Sŏnjong 1083 brother fraternal

Sŏnjong Hŏnjong 1094 elder son hereditary

Hŏnjong Sukchong 1095 uncle usurpation

Sukchong Yejong 1105 son hereditary (backed by a regent)

Yejong Injong 1122 elder son hereditary (backed by a regent)

Injong Ŭijong 1146 elder son hereditary

Ŭijong Myŏngjong 1170 brother military coup military regency

Myŏngjong Sinjong 1197 brother fraternal (backed by a regent)

Sinjong Hŭijong 1204 elder son hereditary (backed by a regent)

Hŭijong Kangjong 1211 uncle elected (backed by a regent)

Kangjong Kojong 1213 son hereditary (backed by a regent)

Kojong Wŏnjong 1259 son hereditary (backed by a regent)

72 Javier Cha

Table 2B. The Background of Early and Middle Koryŏ Queens, 918–1170

King QueenRelation-

shipConsanguinity

DegreeQueen’sFather

QF’sChoronym

Trend

T’aejo 神惠王后 柳氏 1 non-kin N/A Yu Ch’ŏn’gung Chŏngju strongmen

莊和王后 吳氏 2 non-kin N/A O Taryŏn’gun Naju

神明王后 劉氏 3 non-kin N/A Yu Kŭngdal Ch’ungju

神靜王后 皇甫氏

4 non-kin N/A Hwangbo Chegong

Yŏngch’ŏn

神成王后 金氏 5 non-kin N/A Kim Ŏngnyŏm Kyŏngju

貞德王后 柳氏 6 non-kin N/A Yu Tŏgyŏng Chŏngju

Hyejong 義和王后 林氏 1 non-kin N/A Im Hŭi Chinju

Chŏngjong 文恭王后 朴氏 1 non-kin N/A Pak Yŏnggyu Sunch’ŏn

文成王后 朴氏 2 non-kin N/A Pak Yŏnggyu Sunch’ŏn

Kwangjong 大穆王后 皇甫氏

1 (Si) 2 King T’aejo Kaesŏng close kin

Kyŏngjong 獻肅王后 金�